Comparative genomic analysis of bacteria species

National Library Of Medicine

Investigators

Linked publications, trials & patents

Abstract

In the past decade, the amount of microbial genomic data has grown considerably. These enormous datasets present new challenges and opportunities to understand, and even manipulate, the human microbiome for human health. One of the key challenges is that while metagenomic sequence data can be used to identify correlations between microbiome composition and disease states, they should be analyzed in conjunction with experimental approaches to reveal the underlying mechanism for the causal relationship. However, microbiologists have generally focused on model species or a few pathogens when they study a function. Therefore, many functions are known in only one or a few species. In this project, we used comparative genomics and analyses of omics datasets to help identify these functions in different bacterial genomes providing valuable insights into relevance of these functions to human health. To aide in our overall goal of understanding the role of different microbial traits in human health we have developed the ProkFunFind search tool. ProkFunFind is a computational tool that can be used to characterize the presence or absence of different gene systems in microbial genomes. Searches using the ProkFunFind framework focus on identifying collections of genes that represent user defined functions. The identification of these genes is done through an integrated search that can utilize sequence similarity, profile HMMs, protein domains, and orthology definitions to identify putative gene candidates in genomes of interest. This search approach facilitates more sensitive and specific identification of functions through its flexible query format and emphasis on identifying systems of genes related to a function rather than single genes. The development of this tool provides us with a foundation to explore the presence and relevance of different microbial traits in the human microbiome. We have applied the ProkFunFind tool in multiple projects to assist in identifying genes of interest that encode functions that impact human health. A recent application has been for the characterization of a bacterial metabolic pathway for the production of a highly estrogenic compound, S-equol, that has been associated with various positive health effects. Like many functions of interest, equol production had been characterized in a few well studied and cultivated organisms, but the broader distribution of the pathway was not known. Through our analysis we determined that the equol production pathway was encoded by a highly conserved gene cluster found in a small subset of bacterial strains from the Eggerthellaceae family of the Actinobacteria. This limited taxonomic distribution was corroborated through an analysis of a large collection of human gut metagenome datasets. Further analysis of these gene clusters revealed the importance of horizontal gene transfer in the evolution of this pathway. This study provided valuable insights into the structure of the equol gene cluster and its evolutionary history providing a foundation for future studies related to the use of equol producing organisms as probiotics. In addition to our research, we have engaged in multiple collaborative projects related to bacterial and viral genomics. We have continued our collaboration with Dr. Hall (University of Maryland), Dr. Gisela Storz (NICHD/DIR), and Dr. Yigang Tong (Beijing University of Chemical Technology). In the project with Dr. Hall, we analyzed azoreductases in the human gut microbiome revealing their broad distribution and the presence of previously uncharacterized azoreductase enzymes in multiple common gut bacterial strains. In subsequent experiments we confirmed the novel reductases from strains of Fusobacterium, Bacteroides, and Clostridium were functional and suggesting that they may have a previously unrecognized role in drug metabolism in the human gut through azo reduction. In the project with Dr. Gisela Storz, we additionally characterized the Rpn gene and found it was frequently associated with selfish genetic elements and predicted the potential function and interaction between different domains of the Rpn gene, providing insight into how the toxin and antitoxin pairs function together. Lastly, in the project with Dr. Tong, we worked on the development of a new approach for the identification of temperate phages using next-generation sequencing data and through the application of this approach we significantly expanded the catalogue of known temperate phage genomes.